Downloaded from http://www.jci.org on June 17, 2017. https://doi.org/10.1172/JCI5468
Cell-wall determinants of the bactericidal action of group
IIA phospholipase A2 against Gram-positive bacteria
Amy K. Foreman-Wykert,1 Yvette Weinrauch,1 Peter Elsbach,1,2 and Jerrold Weiss3
1Department
of Microbiology, and
of Medicine, New York University School of Medicine, New York, New York 10016, USA
3Department of Internal Medicine and Microbiology, Division of Infectious Diseases,
Inflammation Program, University of Iowa College of Medicine, Iowa City, Iowa 52242, USA
2Department
Address correspondence to: Peter Elsbach, Department of Medicine, New York University School of Medicine,
550 First Avenue, New York, New York 10016, USA. Phone: (212) 263-5633; Fax: (212) 263-8276;
E-mail: [email protected]
Received for publication October 9, 1998, and accepted in revised form January 14, 1999.
We have shown previously that a group IIA phospholipase A2 (PLA2) is responsible for the potent bactericidal activity of inflammatory fluids against many Gram-positive bacteria. To exert its antibacterial
activity, this PLA2 must first bind and traverse the bacterial cell wall to produce the extensive degradation of membrane phospholipids (PL) required for bacterial killing. In this study, we have examined the
properties of the cell-wall that may determine the potency of group IIA PLA2 action. Inhibition of bacterial growth by nutrient deprivation or a bacteriostatic antibiotic reversibly increased bacterial resistance
to PLA2-triggered PL degradation and killing. Conversely, pretreatment of Staphylococcus aureus or Enterococcus faecium with subinhibitory doses of β-lactam antibiotics increased the rate and extent of PL degradation and/or bacterial killing after addition of PLA2. Isogenic wild-type (lyt+) and autolysis-deficient
(lyt–) strains of S. aureus were equally sensitive to the phospholipolytic action of PLA2, but killing and lysis
was much greater in the lyt+ strain. Thus, changes in cell-wall cross-linking and/or autolytic activity can
modulate PLA2 action either by affecting enzyme access to membrane PL or by the coupling of massive
PL degradation to autolysin-dependent killing and bacterial lysis or both. Taken together, these findings
suggest that the bacterial envelope sites engaged in cell growth may represent preferential sites for the
action and cytotoxic consequences of group IIA PLA2 attack against Gram-positive bacteria.
J. Clin. Invest. 103:715–721 (1999)
Introduction
Multicellular organisms developed multiple defense
mechanisms to deal with invasion and infection by
microorganisms. The inflammatory reaction elicited by
the invasion of bacteria mobilizes both cellular (1–4)
and extracellular (5) antimicrobial factors. Using a rabbit model of sterile inflammation, we have shown that
proteins secreted by the cells in the exudate or entering
from the circulation accumulate in the ascitic fluid and
exert potent bactericidal activity against both Grampositive (6) and Gram-negative bacteria (5), as well as
fungi (Weinrauch, Y., and Foreman-Wykert, A.K.,
unpublished observations). A 14-kDa, secretory group
IIA PLA2 present in concentrations ranging from 100 to
1,000 ng/ml is largely responsible for the potent bactericidal activity of these inflammatory fluids against
many Gram-positive bacteria (ref. 6; and Liang, N.S., et
al., manuscript in preparation).
For this PLA2 to exert its bactericidal effects, the enzyme
must first bind to the bacteria and then traverse the multiple peptidoglycan layers of the cell wall to reach and
hydrolyze the phospholipids in the cell membrane. The
thick peptidoglycan layer of Gram-positive bacteria can
be highly cross-linked (7) and may constitute a barrier to
PLA2 access of the cell membrane. In line with this supposition, protoplasts of Staphylococcus aureus and Bacillus
subtilis are more susceptible to PLA2 than intact bacteria
(Liang, N.S., et al., manuscript in preparation). However,
The Journal of Clinical Investigation
|
the potency of mammalian group IIA PLA2 (LD90 1–10
nM) (6) against S. aureus that contain highly cross-linked
cell-wall peptidoglycans (7) implies that, even in these
organisms, regions may exist that permit enzyme penetration to the cell membrane. In this study, we have examined the effect of cell growth, bacterial autolysins, and
cell-wall active antibiotics on the phospholipolytic and
bactericidal activity of PLA2. Our findings suggest that
bacterial envelope sites engaged in the normal cell-wall
turnover associated with cell growth, division, and separation are the sites at which PLA2 preferentially acts
against Gram-positive bacteria.
Methods
Bacteria. Staphylococcus aureus RN450 (8325-4) (8) is a laboratory strain provided by B. Kreiswirth (Public Health Research
Institute, New York, New York, USA). S. aureus Lyt– is an autolysis-deficient, isogenic mutant of strain RN450 (9) and was provided by R.K. Jayaswal (Illinois State University, Normal, Illinois, USA). S. aureus strain 18 and Enterococcus faecium strains 4
and 6 are multi-drug–resistant clinical isolates obtained from
P. Tierno and K. Inglima (Department of Clinical Microbiology, Tisch Hospital, New York, New York, USA). E. faecium strain
4 is highly susceptible to PLA2 (LD50 ∼50 ng/ml), whereas E. faecium strain 6 is less susceptible to PLA2 (LD50 ∼500 ng/ml).
The S. aureus strains were grown overnight at 37°C, washed
once, and then subcultured for either 2.5–3 h (mid-logarithmic
phase) or ∼18 h (stationary phase) in fresh Trypticase Soy Broth
(TSB; Difco Laboratories, Detroit, Michigan, USA) at a starting
March 1999
|
Volume 103
|
Number 5
715
Downloaded from http://www.jci.org on June 17, 2017. https://doi.org/10.1172/JCI5468
Figure 1
Stationary phase and starved S. aureus are less susceptible to PLA2mediated killing and phospholipid degradation. [114C]oleate–labeled S. aureus RN450 (107/ml), harvested from midlogarithmic and stationary phase, were incubated in TSB
supplemented with 1% BSA, 2 mM CaCl2, and 10 mM HEPES (pH
7.4) with or without 250 ng/ml of purified rabbit AF group IIA PLA2
at 37°C. At the indicated times, samples were removed to measure
viability (a) and phospholipid degradation (b) as described in
Methods. Alternatively, [1-14C]oleate–labeled S. aureus RN450
(107/ml) harvested from mid-logarithmic phase were incubated in
fresh or used (nutrient-depleted) TSB supplemented with 1% BSA,
2 mM CaCl2 and 10 mM HEPES (pH 7.4) with and without 150
ng/ml of purified rabbit AF PLA2 at 37°C (c and d). At the indicated times, samples were removed to measure viability (c) and phospholipid degradation (d). The results shown represent the mean ±
SEM of three or more experiments. Some error bars are masked by
the symbols. (open squares) Mid-logarithmic bacteria in TSB; (closed
squares) mid-logarithmic bacteria + PLA2 in TSB; (open circles) stationary bacteria in TSB; (closed circles) stationary bacteria + PLA2 in
TSB; (open triangles) mid-logarithmic bacteria in used TSB; (closed
triangles) mid-logarithmic bacteria + PLA2 in used TSB. AF, ascitic
fluid; PLA2, phospholipase A2; TSB, Trypticase Soy Broth.
concentration of 1.5 × 107 bacteria/ml. The E. faecium strains
were grown overnight (>16 h) in Brain Heart Infusion Broth
(Difco Laboratories) supplemented with 0.5% (vol/vol) heatinactivated horse serum (GIBCO BRL, Gaithersburg, Maryland,
USA). After harvesting, the bacteria were sedimented by centrifugation at 3,000 g for 5 min and resuspended in sterile physiological saline (Baxter Healthcare Corp., Deerfield, Illinois,
USA) to a concentration of 109 bacteria/ml. The bacteria were
used within 30 min of harvesting.
Collection of ascitic fluid, preparation of ascitic fluid filtrate, and
purification of PLA2. Sterile inflammatory exudates were elicited
in New Zealand white rabbits by intraperitoneal injection of
300 ml of sterile physiological saline (Baxter Healthcare Corp.)
supplemented with oyster glycogen (2.5 mg/ml; United States
Biochemical Corp., Cleveland, Ohio, USA). After 16 h, the
inflammatory exudate was collected, and the ascitic fluid (AF)
was prepared as described previously (5, 6). Protein-poor (<1%
of AF protein concentration) AF filtrate was prepared by filtration of the cell-free inflammatory AF in a centrifugal concentrator (Centricon-10; Amicon, Danvers, Massachusetts, USA)
(1600 g for 1 h) at 4°C (6). The AF filtrate was sterile filtered
(0.45 µm; Nalge Nunc International, Rochester, New York,
USA) and stored at –20°C until used. Group IIA PLA2 from rabbit AF was purified as described previously (6, 10, 11).
Assay of PLA2 catalytic activity. The catalytic activity of PLA2 in
various assay media was measured against autoclaved [114C]oleate–labeled Escherichia coli as described previously
(10–12). Arbitrary units of PLA2 activity were measured as
described previously (6, 12).
Recombinant human PLA2 . Recombinant human group IIA
PLA2 was expressed in E. coli and purified as described previously (ref. 13; and Liang, N.S., et al., manuscript in preparation) .
Radiolabeling of S. aureus lipids during growth. To radiolabel the
lipids of S. aureus, the bacteria were subcultured to mid-logarithmic phase in TSB supplemented with 1 µCi/ml of [114C]oleic acid (40–60 mCi/mmol) (Du Pont NEN Research
Products, Boston, Massachusetts, USA) and 0.1% BSA. Stationary phase cultures were grown overnight in TSB supplemented with 2 µCi/ml of [1-14C]oleic acid and 0.1% BSA. Bacteria harvested at mid-logarithmic and stationary phases were
incubated in medium without [1-14C]oleic acid at 37°C for 20
min to chase incorporated, unesterified precursor fatty acid
into ester positions and then were washed as described previ-
716
The Journal of Clinical Investigation
|
ously (6), except that stationary phase bacteria were washed in
medium (TSB) derived from stationary phase cultures (used
TSB). The composition of lipids in bacteria harvested at midlogarithmic or stationary phase was determined by extraction
in CHCl3/CH3OH (14) and analysis by TLC of radiolabeled
material recovered in the CHCl3 phase (>95% of total) (6). In
both mid-logarithmic and stationary phase bacteria, ∼70% of
radiolabeled material was phospholipid, nearly all of which was
phosphatidyl glycerol.
Assay of bacterial phospholipid degradation. Because phospholipid
breakdown products formed during PLA2 treatment are quantitatively recovered in the extracellular medium complexed to albumin, whereas undigested phospholipid remains within the bacterial envelope, phospholipid degradation was routinely
measured as accumulation of radioactive material in the supernatant recovered after sedimentation of the bacteria (11,000 g for
4 min). To confirm that release corresponded quantitatively to
phospholipid degradation, the lipids of S. aureus, with and without PLA2 treatment, were extracted using the method of Bligh
and Dyer (14) and resolved as described previously (6). Phospholipid degradation of the control samples (no PLA2) was ≤5%.
Assay of bacterial viability. To determine the effect of purified
PLA2 on bacterial viability, the colony-forming ability of the
bacteria was measured before and after incubation with various
doses of PLA2. Typical incubation mixtures contained 1–4 × 107
bacteria/ml in the desired medium (AF filtrate, TSB, used TSB,
70% pooled human serum [PHS] or RPMI-1640 [BioWhittaker
Inc., Walkersville, Maryland, USA]) supplemented with 10 mM
HEPES (pH 7.4; 20 mM for filtrate) and 1% (wt/vol) BSA (not
added to serum samples). Calcium chloride (2 mM) was added
to TSB, used TSB, and RPMI incubation mixtures. Incubations
were carried out at 37°C for ≤ 3 h. Used TSB was prepared by
centrifugation (11,000 g for 5 min) and sterile filtration (0.45
µm; Nalge Nunc International) of overnight bacterial cultures
(grown in TSB) to remove bacteria and particulate matter. PHS
was collected from blood of healthy human volunteers and prepared as described previously (15). After incubation, aliquots of
bacterial suspensions were serially diluted in sterile physiological saline and plated in 5 ml of molten (50°C) trypticase soy
agar (Difco Laboratories). Bacterial viability was measured after
18–24 h of incubation at 37°C. When antibiotics were used, the
bacteria were incubated with the desired antibiotic for 15 or 30
min at 37°C before the addition of PLA2. All antibiotics were
March 1999
|
Volume 103
|
Number 5
Downloaded from http://www.jci.org on June 17, 2017. https://doi.org/10.1172/JCI5468
Figure 2
Erythromycin pretreatment reduces the susceptibility of S. aureus to PLA2
[1-14C]oleate–labeled S. aureus strain 18 (107/ml) was preincubated with
and without erythromycin (5 µg/ml) in AF filtrate supplemented with 1%
BSA and 20 mM HEPES (pH 7.4) for 15 min before treatment with 50
ng/ml of purified rabbit AF group IIA PLA2 at 37°C. At the indicated times,
samples were removed to measure viability (a) and phospholipid degradation (b) as described in Methods. The results shown represent the mean ±
SEM of three experiments. Some error bars are masked by the symbols.
(open squares) Bacteria alone; (closed squares) bacteria + PLA2; (open circles)
erythromycin + bacteria; (closed circles) erythromycin + bacteria + PLA2.
pholipids of bacteria harvested from mid-logarithmic
than from stationary phase. More dramatic differences
in PLA2 sensitivity were observed when bacteria harvested from mid-logarithmic phase and exposed to PLA2
were compared in either fresh TSB or used TSB (Fig. 1, c
and d). In the latter medium, the bacteria did not grow
and were neither killed by PLA2 nor suffered appreciable
phospholipid degradation. PLA2 activity toward autoclaved bacteria was the same in fresh and used TSB (data
not shown), indicating that the inhibition of PLA2 antibacterial action in used medium reflected an effect on
the bacteria and not on the PLA2 itself. Thus, these findings suggest that the growth state of the bacteria is an
important determinant of bacterial sensitivity to PLA2.
A bacteriostatic antibiotic also reduces the susceptibility of S.
aureus to PLA2. To further test this hypothesis, we examined the effect of a bacteriostatic antibiotic that promptly inhibits bacterial growth under conditions less likely
to produce the broad metabolic effects that accompany
bacterial starvation or the transition to stationary phase.
For this purpose, we used a multi-drug–resistant clinical
isolate of S. aureus (strain 18) that is growth inhibited,
but not killed, by treatment with erythromycin. In the
absence of erythromycin, this strain of S. aureus was highly susceptible to PLA2; 50 ng/ml of this enzyme was sufficient to cause virtually quantitative degradation of
obtained from the Tisch Hospital Pharmacy. The sensitivity of
RN450 to PLA2 was comparable in all media used.
OD measurements. To monitor the growth or lysis of bacteria
during the course of a viability assay, 500-µl aliquots of assay
samples (4 × 107 bacteria/ml) were removed, placed in a
microcuvette (Fisher Scientific Co., Pittsburgh, Pennsylvania,
USA), and the OD measured at 550 nm on a Beckman DU-30
spectrophotometer.
Gram staining. Gram stains of bacterial samples with and without PLA2 treatment were performed using the Bacto-3-Step
Gram stain kit (Difco Laboratories) according to the manufacturer’s directions.
Results
Stationary phase and starved S. aureus are less susceptible to
PLA2. To determine whether bacterial populations with
different growth rates differed in susceptibility to PLA2,
S. aureus RN450 was harvested from mid-logarithmic
and stationary phases; after resuspension at 107/ml in
TSB, it was tested for PLA2 sensitivity. Under these experimental conditions, bacteria harvested from mid-logarithmic or stationary phase displayed different growth
rates for an extended time (Fig. 1a), providing an experimental setting in which the effects of growth differences
on bacterial sensitivity to PLA2 could be tested. Figure 1
(a and b) shows that purified rabbit group IIA PLA2 produced more rapid and extensive degradation of the phosThe Journal of Clinical Investigation
|
Figure 3
Ampicillin promotes and erythromycin inhibits PLA2-mediated killing and
phospholipid degradation of S. aureus. [1-14C]oleate–labeled S. aureus
strain 18 (107/ml) was treated with ampicillin (1 µg/ml) (open circles),
erythromycin (5 µg/ml) (open triangles), or no antibiotic (open squares) for
15 min in PHS buffered with 10 mM HEPES (pH 7.4) before incubation
with increasing doses of purified rabbit AF group IIA PLA2 (5, 15, and 50
ng/ml) at 37°C. After 90 min, aliquots of the samples were removed to
measure (a) viability (open symbols) and (b) phospholipid degradation
(closed symbols) as described in Methods. The results represent the mean
± SEM of three experiments. Some error bars are masked by the symbols.
March 1999
|
Volume 103
|
Number 5
717
Downloaded from http://www.jci.org on June 17, 2017. https://doi.org/10.1172/JCI5468
Figure 4
Cell-wall active antibiotics can act synergistically or additively with PLA2 to kill E. faecium. E. faecium strains 4 (a–c) and 6 (d–f) (105/ml) were treated
with ampicillin (16 µg/ml; a and d), cephalothin (32 µg/ml; b and e), vancomycin (30 µg/ml; c and f), or no antibiotic for 30 min in PHS buffered
with 10 mM HEPES (pH 7.4) before treatment with 50 ng/ml (strain 4) or 500 ng/ml (strain 6) of purified rabbit AF group IIA PLA2 at 37°C. At the
indicated times, samples were removed to measure viability as described in Methods. The results represent the mean ± SEM of four experiments.
Some error bars are masked by the symbols. (open squares) Bacteria alone; (closed squares) antibiotic + bacteria; (open circles) bacteria + PLA2; (closed circles) antibiotic + bacteria + PLA2.
membrane phospholipids within 30 minutes and >3 logs
killing within 90 minutes (Fig. 2, a and b). In contrast,
pretreatment of strain 18 with erythromycin for 15 minutes before PLA2 exposure greatly retarded the degradation of bacterial membrane phospholipids by PLA2 and
rendered the bacteria largely resistant to the bactericidal
action of PLA2 (Fig. 2, a and b).
Erythromycin had no adverse effect on PLA2 activity
toward autoclaved bacteria (data not shown), indicating
that erythromycin was not directly inhibiting the catalytic activity of PLA2. In addition, erythromycin did not
inhibit the initial binding of PLA2 to the bacteria (data
not shown), an essential first step in the antibacterial
action of mammalian group IIA PLA2 (Liang, N.S., et al.,
manuscript in preparation). Thus, these results support
the hypothesis that bacterial growth and cell division
promote the susceptibility of S. aureus to PLA2-mediated
killing, apparently by facilitating the access of PLA2 to
the phospholipids in the bacterial membrane. To further
explore the possibility that the integrity of the cell wall
is a major variable in the sensitivity of S. aureus to host
PLA2, the effect on PLA2 sensitivity of antibiotics that
interfere with peptidoglycan synthesis was examined.
Ampicillin promotes PLA2 action against S. aureus. The preceding findings suggested that cell-wall turnover associated with cell growth and division may promote the antibacterial action of PLA2. If this is correct, weakening of
the cell-wall structure by antibiotics should also render
the bacteria more susceptible to PLA2. To explore this
possibility, S. aureus strain 18 was treated with ampicillin,
erythromycin, or no antibiotic for 15 minutes before
718
The Journal of Clinical Investigation
|
treatment with increasing concentrations of PLA2. As
shown above, in a different medium, bacteria pretreated
with erythromycin were more resistant to the antibacterial and phospholipolytic effects of PLA2 at all concentrations tested (Fig. 3, a and b). In contrast, bacteria
pretreated with subinhibitory doses of ampicillin were
more susceptible than the untreated bacteria to PLA2
(Fig. 3, a and b). Thus, ampicillin pretreatment of S.
aureus significantly increased bacterial susceptibility to
PLA2-mediated phospholipid degradation at sublethal
concentrations of PLA2 (Fig. 3b).
Antibacterial synergy between cell-wall active antibiotics and
PLA2 against E. faecium. Although pretreatment of S.
aureus with subinhibitory doses of ampicillin appreciably increased the PLA2 degradation of bacterial membrane phospholipids, effects on PLA2-triggered killing
were only modest. Reasoning that bactericidal synergy
might be more pronounced toward Gram-positive bacteria that are inherently more resistant to PLA2, we examined the effects of cell-wall active antibiotics on PLA2triggered killing of two multi-drug–resistant clinical
isolates of E. faecium.
As expected, treatment of these two strains with either
β-lactam antibiotics or vancomycin alone had little or no
effect on bacterial viability (Fig. 4, a–f). However, with
both strains, pretreatment with either ampicillin,
cephalothin, or vancomycin increased bacterial killing
produced when LD50 doses of PLA2 were added (Fig. 4,
a–f). The effects of PLA2 were synergistic with both β-lactams (Fig. 4, a, b, d, and e) and additive with vancomycin
(Fig. 4, c and f). These results indicate that PLA2 and β-lacMarch 1999
|
Volume 103
|
Number 5
Downloaded from http://www.jci.org on June 17, 2017. https://doi.org/10.1172/JCI5468
Figure 5
Autolysis-deficient S. aureus are poorly killed and not lysed by PLA2. [1-14C]oleate–labeled S. aureus RN450 (Lyt+) and Lyt– (4 × 107/ml) were incubated in RPMI-1640 supplemented with 1% BSA, 10 mM HEPES (pH 7.4), and 1 mM CaCl2 with or without 300 ng/ml of purified rabbit AF group IIA
PLA2 at 37°C. At the indicated times, aliquots were removed to measure the phospholipid degradation (a), viability (b), and OD (c) of the samples,
as described in Methods. The results represent the mean ± SEM of three experiments. Some error bars are masked by the symbol. (open squares) RN450
control; (closed squares) RN450 + PLA2; (open circles) Lyt– control; (closed circles) Lyt– + PLA2.
tam antibiotics act synergistically toward both multidrug–resistant S. aureus and E. faecium.
Susceptibility of autolysis-deficient S. aureus to PLA 2. Bacterial cell-wall–degrading enzymes (autolysins) play an
important role in cell-wall morphogenesis that accompanies cell growth, division/separation, and peptidoglycan turnover (16, 17). Thus, these enzymes could
contribute to the greater sensitivity of growing bacteria to PLA2 action by producing localized sites of weakened and thinned peptidoglycan (18, 19). To test this
hypothesis, we compared the sensitivity of S. aureus
RN450 (lyt+) and an isogenic, autolysis-deficient
mutant (Lyt–) to PLA2.
In contrast to the effect of bacterial growth inhibition
(Fig. 1, b and d; Fig. 2b; and Fig. 3b), the reduced
autolysin content of the lyt– RN450 mutant did not
render these bacteria more resistant to the phospholipolytic action of PLA2. Thus, PLA2-triggered phospholipid degradation in RN450 (lyt+) and Lyt– is essentially the same (Fig. 5a). However, the lyt– strain is
much less susceptible to killing by PLA2 (Fig. 5b), and,
in contrast to the wild-type strain, is not lysed during
PLA2 treatment (Fig. 5c). Gram stains confirmed the
much greater lysis of the lyt+ than of the lyt– strain
(data not shown). We conclude that bacterial autolysins
play a major role in PLA2-triggered killing and postkilling lysis of S. aureus.
Discussion
In this study, we have shown that the susceptibility of
Gram-positive bacteria to the bactericidal action of mammalian group IIA PLA2 is profoundly affected by the bacterial growth rate. Thus, conditions that affect cell-wall
structure — cell growth, division, and separation, treatment with cell-wall active antibiotics, and the action of
autolysins — all correlate with bacterial sensitivity to
PLA2. S. aureus was less susceptible to the antibacterial
actions of PLA2 when bacterial growth was slowed by
reaching stationary phase (Fig. 1, a and b), by incubation
in a nutrient-depleted medium (Fig. 1, c and d), or by
antibiotic-induced bacteriostasis (Fig. 2). Conversely, susceptibility to PLA2 was increased when cell-wall active
The Journal of Clinical Investigation
|
antibiotics (Figs. 3 and 4) or endogenous autolysins
weakened envelope structure (Fig. 5).
Bacteria undergo multiple structural and physiological changes during the transition from a growing to a
nongrowing state. Alterations include increased crosslinking and thickening of the peptidoglycan (16, 20),
decreased cell-wall turnover (21), transmembrane potential (22), autolysis (9), and changes in membrane phospholipid composition (23) and protein synthesis (24).
Many of these changes alter the susceptibility of the bacteria to antibacterial agents (20, 25, 26). Therefore, under
our experimental conditions the factor(s) responsible for
the decreased sensitivity of the nongrowing bacteria cannot yet be defined precisely. However, the results of this
study lead us to propose that the structure of the envelope peptidoglycan is a major determinant of the actions
of PLA2 against Gram-positive bacteria.
First, the transition from active bacterial growth to a
nongrowing state is accompanied by increased O-acetylation and thickening of the peptidoglycan layer (16, 20)
and a dramatic increase in resistance of S. aureus to the
antibacterial actions of the enzyme, most likely because
of reduced access to membrane phospholipids. In contrast, the autolytic activity associated with cell-wall biogenesis may account for the high sensitivity of growing
bacteria to PLA2. Second, β-lactam antibiotics, which
reduce the cross-linking of peptidoglycan, also promote
the antibacterial action of PLA2.
However, the ability of the PLA2 to reach and hydrolyze
the bacterial phospholipids is not sufficient to kill the bacteria. The demonstration that PLA2-treated, autolysis-deficient S. aureus undergo as much phospholipolysis as their
wild-type counterparts but largely escape killing and lysis
(Fig. 5) indicates that the phospholipolytic and bactericidal actions of the PLA2 need not be linked; and it implies
that activation of autolytic events is also required for
killing of S. aureus by PLA2.Because the lyt+ and lyt– strains
of S. aureus are equally susceptible to phospholipid degradation by PLA2, differences in their autolytic activity (9)
apparently do not affect the access of PLA2 to the bacterial membrane. The survival of the lyt– strain despite the
nearly quantitative degradation of its membrane phosMarch 1999
|
Volume 103
|
Number 5
719
Downloaded from http://www.jci.org on June 17, 2017. https://doi.org/10.1172/JCI5468
Figure 6
Model for PLA2 action against Gram-positive bacteria.
pholipids further points to remarkable repair capabilities
of the bacteria in the absence of autolytic activity.
The activities of autolysins are tightly regulated (27),
both temporally and spatially, to permit the discrete
action needed for cell-wall biogenesis (16, 18, 19), without causing diffuse cell-wall alterations that would be
harmful to the bacteria. Because the growth rates of the
lyt+ and lyt– strains are comparable, the actual level of
autolytic activity coupled to cell growth processes may be
similar in both strains, despite the large difference in
overall autolytic activity.
Based on these observations, we propose a model for
mammalian group IIA PLA2 action against Gram-positive bacteria (outlined in Fig. 6). First, the PLA 2 must
bind and penetrate the cell-wall, a feature that distinguishes this PLA2 from other closely related, but nonbactericidal, PLA2’s (ref. 6; and Liang, N.S., et al., manuscript in preparation). These initial interactions are
dependent on electrostatic attraction between the very
cationic group IIA PLA2 and anionic bacterial envelope
constituents. One bacterial envelope component that
could be particularly important in this context is lipoteichoic acid (LTA), an abundant anionic membrane glycolipid that extends through the cell-wall to the outer
surface. We speculate that the initial binding of PLA2 to
LTA on the outer surface of the bacteria is followed by
penetration of the PLA2 down the repeating glycerol
phosphate chain of LTA permitting PLA2 to reach the
cell membrane. Binding of PLA2 to LTA might also displace the less cationic autolysins, relieving this negative
constraint on autolysin activity (28) and producing discrete cell-wall changes at sites of PLA2 binding that further promote PLA2 penetration. Penetration of PLA2
may be favored at envelope sites with decreased crosslinking and/or increased active autolysins, e.g., sites of
new cell-wall growth.
When the PLA2 causes massive phospholipid degradation, thereby destabilizing the association of LTA with the
cell membrane, a more diffuse activation of autolysins (in
lyt+ strains) may follow and, consequently, result in irreversible bacterial injury. The activities of cytosolic
autolysins, which are controlled by close association with
membrane phospholipids (29, 30), may also be activated
because of the membrane disruption caused by PLA2. The
decreased autolysin content of the autolysis-deficient
720
The Journal of Clinical Investigation
|
strain of S. aureus may explain why, in this situation, the
bacteria are not lysed, but instead remain largely intact
and able to repair the membrane damage caused by PLA2.
Intrinsic differences in cell-wall peptidoglycan cross-linking and/or autolytic activity among Gram-positive bacterial species may account for different sensitivities to PLA2
(6). For example, B. subtilis, which is exquisitely sensitive to
PLA2, has high autolytic activity (31), whereas less susceptible species display lower autolytic activity or possess fewer
autolysins (32). These differences could be linked to different growth patterns of the bacterial species. It has been
speculated that B. subtilis may undergo random or diffuse
cell-surface extension by the addition of new cell wall at
many sites (33), thereby creating more sites at which PLA2
can cross the cell wall. Conversely, enterococcal species,
which are much less sensitive to PLA2 (6), exhibit patterns
of zonal growth (34), where new cell-wall material is inserted at a single region or zone, resulting in fewer sites for
PLA2 access to the membrane phospholipids.
The demonstration of synergism between sublethal
doses of β-lactam antibiotics and low concentrations of
PLA2 against two drug-resistant enterococcal pathogens
(Fig. 4) that differ in their intrinsic susceptibilities to
PLA2 not only further supports the contention that peptidoglycan structure is an important variable in the antibacterial actions of PLA2, but may also have broader biological implications. These findings suggest that even in
antibiotic-resistant bacteria, discrete changes in the cell
wall caused by subinhibitory doses of β-lactam antibiotics (35, 36) are sufficient to promote access of PLA2 to
the membrane phospholipids, setting in motion both
their degradation and autolytic events resulting in killing.
In contrast to the synergistic effects of β-lactam
antibiotics and PLA2, the combined effects of PLA2 and
vancomycin are additive. This difference is observed
over a wide range of vancomycin concentrations,
including doses at which β-lactam antibiotics and vancomycin, by themselves, have similar effects on bacterial growth and viability (data not shown). The finding
that β-lactam antibiotics, which decrease peptidoglycan cross-linking, acted synergistically with PLA2,
whereas combinations with vancomycin (an agent that
inhibits cell-wall synthesis) were only additive, may
again reflect an important role of the peptidoglycan
structure in PLA2 action.
March 1999
|
Volume 103
|
Number 5
Downloaded from http://www.jci.org on June 17, 2017. https://doi.org/10.1172/JCI5468
In conclusion, this study provides an expansion of our
understanding of the antibacterial action of a group IIA
PLA2. Recently it has become apparent that this enzyme is
an important component of host defense against many
Gram-positive pathogens (6, 37–40). Much effort has been
directed in the past at developing inhibitors of this PLA2
based upon the belief that the products of its action might
exacerbate inflammation. More work is needed to determine the relative prominence of the role of this PLA2 in
antibacterial host defense and the extent of its potential
contribution to harmful or beneficial inflammatory
responses by acting against host lipids. The demonstration in this study of the synergistic and additive actions of
this PLA2 against common pathogens, including multidrug–resistant organisms, raises the prospect of new therapeutic approaches.
Acknowledgments
We gratefully acknowledge the generous help of Barry
Kreiswirth, Radheshyam Jayaswal, and Philip Tierno, and of Ken
Inglima in making bacterial strains and clinical isolates available
for study. We also thank Ning-Sheng Liang for making recombinant human D48S PLA2 available for use, and Jake Kaufman
for technical help. This work was supported by United States
Public Health Service grant AI-18571 and a Ford Foundation
Predoctoral Fellowship (to A.K. Foreman-Wykert).
1. Elsbach, P., and Weiss, J. 1992. Oxygen-independent antimicrobial systems of phagocytes. In Inflammation: basic principles and clinical correlates.
J.I. Gallin, I.M. Goldstein, and R. Snyderman, editors. Raven Press. New
York, NY. 603–636.
2. Klebanoff, S.J. 1992. Oxygen metabolites from phagocytes. In Inflammation: basic principles and clinical correlates. J.I. Gallin, I.M. Goldstein, and
R. Snyderman, editors. Raven Press. New York, NY. 541–588.
3. Ganz, T., and Lehrer, R.I. 1997. Antimicrobial peptides of leukocytes.
Curr. Opin. Hematol. 4:53–58.
4. Levy, O. 1996. Antibiotic proteins of polymorphonuclear leukocytes. Eur.
J. Haematol. 56:263–277.
5. Weinrauch, Y., et al. 1995. Extracellular accumulation of potently microbicidal bactericidal/permeability-increasing protein and p15s in an
evolving sterile rabbit peritoneal inflammatory exudate. J. Clin. Invest.
95:1916–1924.
6. Weinrauch, Y., Elsbach, P., Madsen, L.M., Foreman, A., and Weiss, J. 1996.
The potent anti–Staphylococcus aureus activity of a sterile rabbit inflammatory fluid is due to a 14-kD phospholipase A2. J. Clin. Invest.
97:250–257.
7. Wilkinson, B.J. 1997. Biology. In The staphylococci in human disease. K.B.
Crossley and G.L. Archer, editors. Churchill Livingstone. New York, NY.
1–38.
8. Novick, R. 1967. Properties of a cryptic high-frequency transducing
phage in Staphylococcus aureus. Virology. 33:155–166.
9. Mani, N., Tobin, P., and Jayaswal, R.K. 1993. Isolation and characterization of autolysis-defective mutants of Staphylococcus aureus created by
Tn917-lacZ mutagenesis. J. Bacteriol. 175:1493–1499.
10. Forst, S., et al. 1986. Structural and functional properties of a phospholipase A2 purified from an inflammatory exudate. Biochemistry.
25:8381–8385.
11. Wright, G.W., Ooi, C.E., Weiss, J., and Elsbach, P. 1990. Purification of a
cellular (granulocyte) and an extracellular (serum) phospholipase A2
that participate in the destruction of Escherichia coli in a rabbit inflammatory exudate. J. Biol. Chem. 265:6675–6681.
12. Elsbach, P., and Weiss, J. 1991. Utilization of labeled Escherichia coli as
phospholipase substrate. Methods Enzymol. 197:24–31.
13. Weiss, J., Inada, M., Elsbach, P., and Crowl, R.M. 1994. Structural determinants of the action against Escherichia coli of a human inflammatory
fluid phospholipase A2 in concert with polymorphonuclear leukocytes.
J. Biol. Chem. 269:26331–26337.
14. Bligh, E.S., and Dyer, W.J. 1959. A rapid method of total lipid extraction
and purification. Can. J. Biochem. Physiol. 37:911–917.
The Journal of Clinical Investigation
|
15. Madsen, L.M., Inada, M., and Weiss, J. 1996. Determinants of activation
by complement of group II phospholipase A2 acting against Escherichia
coli. Infect. Immun. 64:2425–2430.
16. Wong, W., Chatterjee, A.N., and Young, F.E. 1978. Regulation of bacterial cell walls: correlation between autolytic activity and cell wall turnover
in Staphylococcus aureus. J. Bacteriol. 134:555–561.
17. Sugai, M., et al. 1995. Identification of endo-beta-N-acetylglucosaminidase and N-acetylmuramyl-L-alanine amidase as cluster-dispersing enzymes in Staphylococcus aureus. J. Bacteriol. 177:1491–1496.
18. Sugai, M., et al. 1997. Localized perforation of the cell wall by a major
autolysin: atl gene products and the onset of penicillin-induced lysis of
Staphylococcus aureus. J. Bacteriol. 179:2958–2962.
19. Yamada, S., et al. 1996. An autolysin ring associated with cell separation
of Staphylococcus aureus. J. Bacteriol. 178:1565–1571.
20. Johannsen, L., Labischinski, H., Reinicke, B. and Giesbrecht, P. 1983.
Changes in the chemical structure of walls of Staphylococcus aureus grown
in the presence of chloramphenicol. FEMS Microbiol. Lett. 16:313–316.
21. Boothby, D., et al. 1973. Turnover of bacterial cell wall peptidoglycans. J.
Biol. Chem. 248:2161–2169.
22. Koo, S.P., Bayer, A.S., Sahl, H.G., Proctor, R.A., and Yeaman, M.R. 1996.
Staphylocidal action of thrombin-induced platelet microbicidal protein
is not solely dependent on transmembrane potential. Infect. Immun.
64:1070–1074.
23. Short, S.A., and White, D.C. 1971. Metabolism of phosphatidylglycerol,
lysylphosphatidylglycerol, and cardiolipin of Staphylococcus aureus. J. Bacteriol. 108:219–226.
24. Watson, S.P., Clements, M.O., and Foster, S.J. 1998. Characterization of
the starvation-survival response of Staphylococcus aureus. J. Bacteriol.
180:1750–1758.
25. Tuomanen, E., Cozens, R., Tosch, W., Zak, O., and Tomasz, A. 1986. The
rate of killing of Escherichia coli by beta-lactam antibiotics is strictly proportional to the rate of bacterial growth. J. Gen. Microbiol. 132:1297–1304.
26. Koo, S.P., Yeaman, M.R., and Bayer, A.S. 1996. Staphylocidal action of
thrombin-induced platelet microbicidal protein is influenced by
microenvironment and target cell growth phase. Infect. Immun.
64:3758–3764.
27. Snowden, M.A., and Perkins, H.R. 1991. Cross-linking and O-acetylation
of peptidoglycan in Staphylococcus aureus (strains H and MR-1) grown in
the presence of sub-growth-inhibitory concentrations of beta-lactam
antibiotics. J. Gen. Microbiol. 137:1661–1666.
28. Qoronfleh, M.W., and Wilkinson, B.J. 1986. Effects of growth of methicillin-resistant and -susceptible Staphylococcus aureus in the presence of
beta-lactams on peptidoglycan structure and susceptibility to lytic
enzymes. Antimicrob. Agents Chemother. 29:250–257.
29. Brunskill, E.W., and Bayles, K.W. 1996. Identification and molecular
characterization of a putative regulatory locus that affects autolysis in
Staphylococcus aureus. J. Bacteriol. 178:611–618.
30. Fischer, W. 1994. Lipoteichoic acid and lipids in the membrane of Staphylococcus aureus. Med. Microbiol. Immunol. 183:61–76.
31. Cleveland, R.F., Daneo-Moore, L., Wicken, J.A., and Shockman, G.D.
1976. Effect of lipoteichoic acid and lipids on lysis of intact cells of Streptococcus faecalis. J. Bacteriol. 127:1582–1584.
32. Cleveland, R.F., Wicken, A.J., Daneo-Moore, L., and Shockman, G.D.
1976. Inhibition of wall autolysis in Streptococcus faecalis by lipoteichoic
acid and lipids. J. Bacteriol. 126:192–197.
33. Foster, S.J. 1992. Analysis of the autolysins of Bacillus subtilis 168 during
vegetative growth and differentiation by using renaturing polyacrylamide gel electrophoresis. J. Bacteriol. 174:464–470.
34. Rogers H.J., Perkins, H.R., and Ward, J.B. 1980. Microbial cell walls and
membranes. In The bacterial autolysins. Chapman and Hall. London, United Kingdom. 437–459.
35. Doyle, R.J., and Koch, A.L. 1987. The functions of autolysins in the
growth and division of Bacillus subtilis. Crit. Rev. Microbiol. 15:169–222.
36. Higgins, M.L., and Shockman, G.D. 1970. Model for cell wall growth of
Streptococcus faecalis. J. Bacteriol. 101:643–648.
37. Harwig, S., et al. 1995. Bactericidal properties of murine intestinal phospholipase A2. J. Clin. Invest. 95:603–610.
38. Qu, X.D., Lloyd, K.C., Walsh, J.H., and Lehrer, R.I. 1996. Secretion of type
II phospholipase A2 and cryptdin by rat small intestinal Paneth cells.
Infect. Immun. 64:5161–5165.
39. Qu, X.D., and Lehrer, R.I. 1998. Secretory phospholipase A2 is the principal bactericide for staphylococci and other gram-positive bacteria in
human tears. Infect. Immun. 66:2791–2797.
40. Weinrauch, Y., Abad, C., Liang, N.-S., Lowry, S.F., and Weiss, J. 1998.
Mobilization of potent plasma bactericidal activity during systemic bacterial challenge: role of group IIA phospholipiase A2. J. Clin. Invest.
102:633–638.
March 1999
|
Volume 103
|
Number 5
721